| Photo by Jane Scherr |
Roy Caldwell Interview: Conversations with History; Institute of International Studies, UC Berkeley
| Photo by Jane Scherr |
Page 6 of 8
Now let's go back to the morphology of this creature. One of the things that you wound up focusing on as you explored this creature is vision. As part of the development of these tools for aggression, you were led to explore its capabilities for seeing. Tell us about that and the extraordinary complexity this creature manifests.
Many of the animals, particularly the smashers, which have this dangerous weapon system, are extremely brightly colored. They use those colors in a variety of displays where they threaten with their appendages and they jump up and down and flare things up and such. And you can't look at them and not think, at least anthropomorphically -- they're using these color displays -- "Well, if they're using these brightly colored displays, they probably are seeing color."
As luck would have it, I was at a meeting in Italy several years ago, and I met another biologist named Tom Cronin, who worked on sensory physiology of invertebrates, particularly crustaceans. We started talking about stomatopods and color vision and decided maybe we should look at that.
Then we hooked up with another biologist, a guy named Justin Marshall from, at that time, Sussex, and the three of us started trying to figure out, did they see color? That's not an easy question to figure out, actually. We tried showing them color targets and asking them, can you tell the green square from the red square? They didn't always want to tell us the answer.
Eventually, I worked out a technique. We took a little plastic cube and drilled a hole in the center of it, and then we put some shrimp in that hole. Then we glued a piece of glass over the top of the hole. If the animal wanted to get the food, it had to break the glass, which it could do, and then pull it out. It was like opening a snail. And then we put different pieces of colored plastic on the cube. We'd give the animal, say, a blue one and a green one. And it would be trained to open the green one for food. And they learned it very quickly. So, that started to tell us that they can see color. At the same time, Justin and Tom started looking at the morphology of the eye and the histology of the eye, using some very sophisticated techniques. They discovered first of all that there were filters, colored filters, inside the eye.
Not in all of the eye; just in six rows of visual elements that go down and bisect the eye.
These color filters were stacked up one on top of another, and it reminded us very much of some of the early satellites, the remote sensing satellites that could detect color on earth by using a black and white vitacon tube. You take a picture through one filter and then you take a picture through another filter and then you use a computer to figure out the color image. They have that capability, it appeared. And that was our first idea of how they were doing it.
Then we started looking at the visual pigments. We see color with three primary visual pigments. It turned out the stomatopods have up to sixteen, that have peaks of spectral sensitivity ranging from about 350 nanometers, well under the UV, way beyond what we can see, out to about 700. So, they have tremendous capacity in terms of the visual pigments, and then you add on top of that the color filters, and that package has a mind-boggling capability of distinguishing spectral differences.
Some of our most recent work has now shown that that system is tunable. Let's say we have an animal that lives from the surface down to 100 feet. At the surface, there's a broad spectrum of light ranging from UV out to red, and in fact some of those filters go all the way from transmitting light in the red down to the blue. You take the same animal and put it at 100 feet, there's only basically blue light. And if it stays there for a few weeks, the filters change color and transmit different wavelengths. They narrow or tune the light that's transmitted more towards the blue. It makes the eye much more sensitive to light, but it sees much less diversity of the spectrum. So, it's a pretty complicated eye.
As I mentioned, they can see polarized light. There's also the capability of incredible range-finding capacity. We can tell how far something away is by parallax. We have two eyes and the images are slightly different. Stomatopods actually have three different focal points in one eye, three different focal points in the other eye, so it has hexnocular vision to tell how far away things are. So, it's a very, very complicated eye that works very differently than the way ours does. Like I said, there is this row of ommatidia, or visual units that come down over the eye. All of the color and the UV and the polarizing capacity are within that row, so it scans the world by moving the eye stalk around and rotating it. So it's sort of like a scanner looking at the world. When it gets really interested in something, the two eyes come over together and the rows actually cross.
So it's doing this line-by-line.
It's basically line-by-line, yes.
In a way, we've started with the morphology again, that is, that the sight, in part, comes with this armature and its function as an aggressor, on the one hand, but it's also responding to its environment, because the ocean is a place where the range of light must be quite extraordinary.
Yes. From the surface, the same light that we're seeing right now. By the time you get to 100 or 120 feet it's just blue. A lot of UV and polarized light in this environment, much less when you get deeper.
Now, in terms of the importance of vision for the animal's needs, it must be related to dealing with prey, getting food. It must also be related to mating, right?
Yes. You have to be able to find a member of the opposite sex of your species. You have to be able to find animals that you've already interacted with. "Have I beaten this guy up before or can it beat me up?" So, it's information about your competitors. "Is that fish a parrot fish which is just going to munch on coral, or is that a jack, which is going to eat me? Is that a blue-ringed octopus which might eat me, or just a piece of algae?" There are a lot of things that are color-coded. In fact, if you go into a reef in the Pacific you may find five or six species of mantis shrimp all living in the same habitat using the same type of holes, and they'll be color-coded. One will have a big pink spot here, another one an orange, another one a blue, another one a white. So vision and color is very important to them.
The question is why the eye is so complex, compared to, say, our eye, which has these three visual figments that are taking in three different spectra; and then our brain is taking that information and generating a color image and then figuring out depth and texture and all that other stuff.
So there is a lot of complexity in the human brain for filtering the information.
About a fifth or a sixth of our brain back here is involved in integrating, extracting information from what comes from the eye. A stomatopod has a little tiny brain, actually, and apparently vision is very important for it, and so what it has done is process visual information peripherally by specific receptor types, and then that information is fed straight into the brain to be used, whereas we send sort of general information in and then process it centrally.
You mentioned earlier when we were talking that you came through experiments to the conclusion that these creatures had memory, basically, and that this was a system that could be useful in remembering a previous encounter. Tell us about that.
[This is] one of these serendipitous discoveries that comes through the stupidity of an investigator. I was looking at aggressive behavior and fighting in these animals and trying to figure out what determined who would win. And in putting the animals together, one day I made a mistake I put animals together that had met previously, about a week earlier, and they didn't fight. They just sort of looked at each other and went their separate ways.
But they had fought in the previous encounter?
Yes, they had fought before. And I thought, "Wait a minute, maybe they knew one another, and they remembered one another." So I designed a very simple experiment based on odor. I was very quickly able to show that they could identify the odor of a previous opponent and remember whether they had won or lost against that individual. And in fact, we later showed that mates can remember each other by odor for up to a month and adjust their aggressive behavior accordingly. They have a very long-term memory, true individual recognition, which is almost unheard of in invertebrates.
Next page: The Aquarius Mission
© Copyright 2001, Regents of the University of California